Oxygen isotopic compositions of the tests of planktonic foraminifera from several Deep Sea Drilling Project sites provide a general picture of low-latitude marine temperatures from Maastrichtian time to the present. Bottom temperatures determined from the isotopic compositions of benthonic foraminifera are interpreted as being indicative of high-latitude surface temperatures. Prior to the beginning of middle Miocene time, high- and low-latitude temperatures changed in parallel fashion. Following an apparently small and short-lived drop in temperature near the Tertiary-Cretaceous boundary, temperatures remained warm and relatively constant through Paleocene and early and middle Eocene time; bottom temperatures then were on the order of 12°C. A sharp temperature drop in late Eocene time was followed by a more gradual lowering of temperature, culminating in a late Oligocene high-latitude temperature minimum of about 4°C. A temperature rise through early Miocene time was followed in middle Miocene time by a sudden divergence of high- and low-latitude temperatures: high-latitude temperatures dropped dramatically, perhaps corresponding to the onset of major glaciation in Antarctica, but low-latitude temperatures remained constant or perhaps increased. This uncoupling of high-and low-latitude temperatures is postulated to be related to the establishment of a circum-Antarctic circulation similar to that of today. A further drop in high-latitude temperatures in late Pliocene time probably signaled the onset of a major increase in polar glaciation, including extensive sea-ice formation.
Early Miocene, small-amplitude (1 per mil) sympathetic fluctuations in isotopic compositions of planktonic and benthonic foraminifera have been identified. These have a period of several hundred thousand years. Superimposed upon these are much more rapid and smaller fluctuations (0.2 to 0.5 per mil) with a period of about 80,000 to 90,000 yr. This is similar to the period observed for Pleistocene isotopic temperature fluctuations.
In low latitudes, much smaller vertical temperature gradients seem to have existed during Maastrichtian and Paleogene time than exist at present. The absence of a sharply defined thermocline during early Tertiary time is also suggested.